The Swap Spread Curve and Box Trades

A single swap spread — the gap between a swap rate and a Treasury yield of the same maturity — reflects one point in time on one maturity. But swap spreads at 2 years, 5 years, 10 years, and 30 years are driven by overlapping but distinct forces: repo scarcity and short-end supply dominate at 2 years, while pension and insurance demand for long-dated fixed receiving dominates at 30 years. Because those drivers diverge, the swap-spread curve itself has a shape, and that shape moves. Trading two points on that curve against each other — long the spread at one maturity, short it at another — is called a box trade, because on a diagram of swap rates versus Treasury yields, the four legs trace out a rectangle.

Building the box

A 5s30s swap spread box combines four legs: long the 5-year Treasury, receive fixed on a 5-year swap, short the 30-year Treasury, pay fixed on a 30-year swap (sized to be duration neutral on each leg). Equivalently, it's "long the 5-year swap spread, short the 30-year swap spread." If both spreads happen to move by the same amount in the same direction, driven by some common shock like a broad flight to swaps, the trade is flat — that common move cancels. The position only makes or loses money on the difference between the two spreads, which is exactly the point: it strips out the swap-spread level and isolates its curve shape.

Worked example

The 5-year swap spread is currently -10bp (swaps yield 10bp less than Treasuries) and the 30-year swap spread is -45bp. A trader believes long-end swap spreads are unusually depressed due to heavy pension receiving demand and will normalize (become less negative) faster than the 5-year spread, so they put on a box: long the 5s spread, short the 30s spread, sized DV01-neutral on each leg. Three months later, the 5-year spread is still -10bp but the 30-year spread has moved to -30bp (up 15bp). The relative move the trader was betting on, +15bp on the box (30y spread improving relative to 5y), plays out even though neither Treasury yield nor either swap rate individually needed to move in a predictable way — only the gap between the two spreads had to close as expected.

What this means in practice

Box trades let relative-value desks express views on structural, slow-moving forces — pension fund receiving flows at the long end, dealer balance-sheet costs at the short end, Treasury issuance skewed toward certain tenors — without taking a view on the outright level of rates or even the outright level of swap spreads. They're popular exactly because level-of-rates risk is hard to forecast and manage, while relative curve-shape dislocations driven by known structural flows are comparatively easier to reason about and size.